Several series of experiments are proposed in a continuing effort to understand the neural mechanisms that mediate visual detection. The different series represent three different empirical approaches to the problem of luminance detection at threshold. In Series 1, neurophysiological recordings from single ganglion cell axons in the optic tract of the goldfish will be compared mathematically with the animal's psychophysical frequency-of-seeing curve, in an attempt to provide the first direct (within species) evidence for the nature of the transforms between the behavior of ganglion cells and the probability of producing a particular psychophysical response. Some experiments in this series will involve testing fish of different sizes; because of the differential rates of addition of rods and cones to this animal's retina throughout life, the hypothesis that relative rod- and cone-mediated sensitivity might shift with age is tenable, and the outcome of this investigation may provide new information about the determinants f visual sensitivity via both scotopic and photopic pathways. Experiments in Series 2 are intended to provide the first complete description of cyclic fluctuations in visual sensitivity in two vertebrate species, goldfish and human; preliminary evidence suggests that such changes do exist in both species. I propose to record absolute psychophysical scotopic and photopic and photopic detection threshold in both species at 3 hr intervals throughout the 24 hr day, during all seasons of the year. Additional experiments are planned to determine whether the rhythm is circadian and whether it originates within or beyond the eye. Series 3 will provide a first inquiry into the neural mechanisms that might mediate the daily changes in visual sensitivity described in Series 2. I propose to examine the responses of goldfish retinal ganglion cells to determine whether some aspect of their pattern of discharge changes systematically with time of day, and if so, whether the removal of the pineal gland affects that rhythm. The successful completion of these three series of experiments will help to increase our knowledge about the relation between the interactions among nerve cells in vertebrate species and the ability to detect light.